In general, techniques for synthesizing images desired by a user on a network are widely known.
Synthesized image is usually generated by overlapping and joining a number of pre-existing images such as images having already been stored in a server connected through a network or images having been made available from digital still cameras. In some cases, the synthesized image may be generated by a texture mapping.
However, there are problems in that synthesized images generated by using the above conventional methods give rise to the occurrences of motion blur. For example, motion blur commonly occurs with a moving body that cannot be accurately adjusted into the synthesized images. Consequently, synthesizing processes are performed with low accuracy and unnatural appearances are frequently generated in the synthesized images.
Additionally, if one server performs synthesizing process in a multiple bundle task, a significant duration of time is required for the process. For example, even when the same background images are to be overlapped, the overlapping process is repeated as many times as the number of images desired. This poses a problem to users in that when the processing services are charged by time, the cost can be a considerable burden.